1. Field of the Invention
The present invention relates to a detection field, and more specifically, to a system detection method and an apparatus for system detection, and a flow control method and a device for flow control.
2. Description of Related Art
With the development of technology, systems composed of devices that process user requests on the server side become more complex and larger. As the number of user requests processed by the system increases, when there is a bottleneck in some resources of the system, response time to request related to these resources will become very large, so a user has to wait for a very long time to get the response and user experience is very poor. Many resources in the system will bottleneck, such as, CPU, storage, input/output etc. all of which can bottleneck, resulting in a sharp decline of system performance.
Thus, as for how to detect system bottleneck and optimize the system, a white box test scheme has been proposed in prior art. This scheme will determine in advance one or more resources that can cause bottlenecks in the system and employ an optimization method for this resource bottleneck, or, this scheme needs to set a monitor for each type of resources in the system and enforces corresponding optimization scheme based on monitor's feedback of whether there is a bottleneck in that resource. Such a scheme has certain problems, for example, as to the one or more predetermined resources that can cause a bottleneck in the system, since there can be one or more resources causing a bottleneck in the system and the number can change dynamically. This predetermined optimization scheme functions only when resources causing a bottleneck in the system are just the predetermined ones and can not function in other cases. If a scheme of setting a monitor for each resource is employed, there can be a problem that the number of required monitors is too many and the optimization scheme is too complex to be actually implemented. Thus, in practice, using white box technology is difficult to achieve effective detection on system bottleneck.
As for TCP, there is a black box detection technology in prior art, which judges whether there is any band width bottleneck through parameters such as packet loss rate and packet disordering, however, this technology is primarily intended for transport layer and is difficult to be applied in a system handling request of user devices. Further, there is no such parameter as packet loss rate in the system, and such black box detection can not be performed.